Evaluation of No Evidence of Disease Activity in a 7-Year Longitudinal Multiple Sclerosis Cohort

Importance  With multiple and increasingly effective therapies for relapsing forms of multiple sclerosis (MS), disease-free status or no evidence of disease activity (NEDA) has become a treatment goal and a new outcome measure. However, the persistence of NEDA over time and its predictive power for long-term prognosis are unknown.

Objective  To investigate NEDA during 7 years as measured by relapses, disability progression, and yearly magnetic resonance imaging (MRI).

Design, Setting, and Participants  Patients were selected from the 2200-patient Comprehensive Longitudinal Investigation of Multiple Sclerosis at Brigham and Women’s Hospital (CLIMB) cohort study. Patients were required to have an initial diagnosis of clinically isolated syndrome or relapsing-remitting MS and a minimum of 7 years of prospective follow-up that included yearly brain MRI and biannual clinical visits (n = 219). Patients were analyzed independent of disease-modifying therapy. Patients were classified as having early (recent-onset) MS if they were 5 years or less from their first MS symptom at enrollment or otherwise considered to have established MS (>5 years from onset).

Main Outcomes and Measures  NEDA was defined as a composite that consisted of absence of relapses, no sustained Expanded Disability Status Scale score progression, and no new or enlarging T2 or T1 gadolinium-enhancing lesions on annual MRI. Relapses, progression, and MRI changes were also investigated as individual outcomes.

Results  A total of 99 of 215 patients (46.0%) had NEDA for clinical and MRI measures at 1 year, but only 17 of 216 (7.9%) maintained NEDA status after 7 years. No differences were found in NEDA status between patients with early vs established MS. A dissociation was found between clinical and MRI disease activity. Each year, 30.6% (64 of 209) to 42.9% (93 of 217) of the cohort had evidence of either clinical or MRI disease activity but not both. NEDA at 2 years had a positive predictive value of 78.3% for no progression (Expanded Disability Status Scale score change ≤0.5) at 7 years. Only minor improvement was found in the positive predictive values with additional follow-up of 1 to 3 years.

Conclusions and Relevance  NEDA is difficult to sustain long term even with treatment. NEDA status at 2 years may be optimal in terms of prognostic value in the longer term. Our results provide a basis for investigating NEDA as an outcome measure and treatment goal and for evaluating the effect of new MS drugs on NEDA.

With multiple and increasingly effective US Food and Drug Administration–approved therapies for relapsing forms of multiple sclerosis (MS), disease-free status or no evidence of disease activity (NEDA) has become a goal for the treatment of MS and a new outcome measure.^1-3Quiz Ref IDNEDA is an outcome measure that represents the most stringent standard of therapeutic efficacy and suggests complete remission of disease. The concept of NEDA has been applied to other conditions, including cancer and infectious diseases, for which treatment may indeed render the patient free of disease. More recently, the disease-free concept has been applied to autoimmune diseases, such as rheumatoid arthritis, after the introduction of disease-modifying therapies, such as anti–tumor necrosis factor drugs. For MS, the term disease-free status has been replaced by NEDA because of the limits of our ability to evaluate the full extent of underlying disease activity.

NEDA has become an important secondary outcome measure in clinical trials with the introduction of more effective types of disease-modifying therapy for relapsing forms of disease.^2,4-10Quiz Ref IDNEDA is defined as the absence of new or enlarging T2 lesions or T1 gadolinium-enhancing lesions on magnetic resonance imaging (MRI) and no sustained Expanded Disability Status Scale (EDSS) score progression or clinical relapse.^2,7 Other definitions have been used, including evaluation of new MRI lesions, relapses, and progression individually or combining the last 2 categories to rule that there is no evidence of clinical disease activity.⁴ The interval for analyzing NEDA or disease-free status has varied from 24 weeks to 3 years depending on the duration and time points of the clinical trial being analyzed.^4-10 Investigators have not discriminated between patients who are early in their disease course vs those with more established disease despite the well-established pattern of more inflammatory disease at onset. It is not known what proportion of patients with MS can be expected to maintain NEDA over time and the relative contribution of clinical vs radiographic disease activity to the loss of NEDA.

The primary aim of our study was to investigate how NEDA evolves during 7 years in a cohort of patients with relapsing-remitting MS (RRMS) who underwent clinical examinations every 6 months and MRI yearly. To better assess the value of NEDA as a useful outcome measure in clinical care and clinical trials, we investigated the minimum interval for assessment of NEDA, which provided reasonable prognostic value for long-term disability. For this, we calculated the predictive value of NEDA annually for absence of disability progression at the end of follow-up. We evaluated the performance on NEDA in patients with early RRMS vs patients with more established disease because inflammatory activity usually wanes with time. Although we focused our present investigation on patients with 7 years of follow-up, the ongoing 2200-patient Comprehensive Longitudinal Investigation of Multiple Sclerosis at Brigham and Women’s Hospital (CLIMB) cohort study will provide future information concerning NEDA in a cohort treated with more aggressive therapies and in patients followed up for longer than 7 years.

The study was approved by the institutional review board at the Brigham and Women’s Hospital. All patients provided written informed consent. Because of intermittent missing clinical visits or MRIs, fewer than 219 patients contributed to the calculation of disease-free status for some of the outcome measures and time points, but at least 207 patients contributed to each calculation.

We studied 219 patients aged 18 to 65 years who were enrolled in the Partners Multiple Sclerosis Center CLIMB study from January 1, 2000, through December 31, 2005, who had at least 7 years of follow-up¹¹ and had a diagnosis of clinically isolated syndrome or RRMS as defined by the 2001 criteria of McDonald et al.¹² Relapses and EDSS scores were evaluated at 6-month intervals during the time of clinic visits. Brain MRI (1.5 T) was performed annually. Spinal cord imaging was available for 162 patients (74.0%). The annual attrition in the CLIMB study is 2% to 5%. The demographic characteristics of the 219-patient cohort are given in Table 1. Patients were classified as having early (recent-onset) MS if 5 years or less had elapsed from their first MS symptom at enrollment and as having established MS if more than 5 years had elapsed from onset.

We used the following definitions. Relapse was defined as the appearance of new symptoms or signs that lasted more than 24 hours without concurrent fever or illness. Relapses were recorded by the treating physician at the face-to-face biannual visits. Progression was defined as an EDSS score increase of 1 or more recorded at a biannual clinical visit that was sustained at the subsequent clinical visit 6 months later. If the EDSS score was 0 at baseline, progression was defined as an EDSS score change of 1.5 or more that was sustained at the subsequent clinical visit. Magnetic resonance imaging activity was defined as new or enlarging T2 hyperintense lesions or T1 gadolinium-enhancing lesions on brain or spinal cord MRI. To qualify as no evidence of MRI activity, new T2 hyperintense lesions and T1 gadolinium-enhancing lesions had to be absent on brain and spinal cord MRIs. Timed 25-ft walk (T25FW) change was defined as an increase of more than 20% on the T25FW relative to the baseline measurement. Because we did not routinely collect T25FW data at the beginning of the study, T25FW data were available for only 145 patients (66.2%).

The primary analysis was the proportion of patients who met NEDA criteria: no relapses, no sustained progression, and no MRI activity during the 7-year study period. Not all 219 patients contributed to each time point as there was an occasional missed MRI or clinical visit. The demographic characteristics of patients classified as having NEDA at year 2 and year 7 were compared with those of patients classified as having evidence of disease using the Fisher exact test for dichotomous outcomes and a t test for continuous outcomes. To assess the association between clinical and radiologic disease activity, we calculated the proportions of patients with clinical disease activity and no MRI disease activity and those with no clinical disease activity but evidence of MRI disease activity. The positive predictive values (PPVs) and negative predictive values (NPVs) of NEDA and each component of NEDA for predicting the absence of progression (EDSS score change ≤0.5) at 7 years were calculated each year. We also calculated the PPV and NPV of NEDA and each component measure to predict lack of change in the T25FW. In addition to the analyses in the whole cohort, we separately evaluated patients with early MS vs those with established MS, and the proportions of patients in these 2 groups who met each definition were compared with the Fisher exact test. All statistical analyses were conducted using the application program R (http://www.r-project.org).

We found that 99 of 215 patients (46.0%) in our cohort met the NEDA criteria at 1 year (no clinical or MRI disease activity). At 2 years, 60 of 218 (27.5%) maintained NEDA status, whereas only 17 of 216 (7.9%) had NEDA at 7 years (Figure 1). Figure 1 also presents the proportion of disease-free patients for individual measures (relapse, progression, relapse and progression, MRI, and T25FW). After 2 years, the rate of loss of NEDA status decreased for all measures except sustained progression, although a plateau was never reached, and there was continued attrition across all measures up to 7 years. The measure that displayed the least change during the 7-year period was progression: 117 of 207 (56.5%) had no evidence of progression at 7 years. Fewer subjects maintained no evidence of MRI activity after the first year (133 out of 215 [61.9%]) compared with no evidence of clinical activity (151 out of 217 [69.6%] for relapse and progression), but a similar proportion of patients were free of MRI disease activity at 7 years (52 of 215 [24.2%]) as the proportion free of clinical disease activity (49 of 213 [23.0%]). When the characteristics of patients classified as having NEDA were compared with those with evidence of disease at 2 and 7 years (Table 1), patients who had NEDA at 2 and 7 years were more likely to be white, to be older, to be diagnosed as having RRMS vs clinically isolated syndrome at enrollment, and to have longer disease duration at baseline, but these differences were minor and not statistically significant (P > .05 for each comparison). When the proportion of patients who developed MRI disease activity after their baseline visit were analyzed, 7.8% to 12.4% each year were classified as developing MRI disease activity solely on the basis of new spinal cord lesions. Although not part of our formal definition of NEDA, we investigated changes in the T25FW and how it was linked to NEDA because the T25FW is a standardized measure used in clinical trials and clinical practice. Studies^13-15 indicate that a 20% change in the T25FW corresponds to a meaningful clinical change that is not expected from routine fluctuation in the T25FW in stable patients. We found that the proportion of patients who developed disease activity based on the T25FW over the 7 years was greater than the proportion based on progression and less than the proportion based on relapses or MRI.

A dissociation was found between absence of disease activity as classified by imaging vs clinical measures (Figure 2). The proportion of patients who had disease activity on one measure but not the other ranged from 42.9% at year 2 to 30.6% at year 7. Quiz Ref IDThe proportion of patients who had no evidence of MRI disease activity but had evidence of clinical disease activity remained relatively constant over time (15.0% in year 1 and 15.8% in year 7). The percentage of patients who had no evidence of clinical disease activity but had evidence of MRI disease activity decreased from 23.5% at year 1 to 14.8% at year 7.

We then investigated the ability of NEDA status at each year to predict change in disability at 7 years. We found that the PPV of NEDA to predict no progression (EDSS score change ≤0.5) at 7 years was 71.7% at 1 year and 78.3% at 2 years. There was minor additional gain in predictive power by evaluating NEDA at years 3 to 5 (maximum PPV of 85.3% at year 4). The PPV of NEDA at 6 years was 95.0% and, by definition, was 100% at 7 years. The NPV of NEDA was relatively constant during the 7-year period, with a minimum NPV of 40.7% and a maximum NPV of 43.1%, which occurred at 2 years. The PPV of the combined outcome measures that comprised NEDA at 2 years was greater than the PPV of the individual outcomes (relapses, progression, MRI disease activity, or clinical disease activity [relapses or progression]) at 2 years. The PPV of the individual NEDA components did not exceed 78.3% at any time point up to 6 years, demonstrating the value of the combined definition. When the change in the T25FW at 7 years was used as the disability outcome instead of the EDSS score, the PPV improved (range, 75.0%-100%) but the NPV was worse (range, 22.0%-27.2%).

A total of 127 patients were classified as having early MS and 92 as having established MS. For the established group, the PPV of NEDA for absence of progression (EDSS score change ≤0.5) at 7 years increased steadily until year 4 and then plateaued until year 6. This trend was similar to that observed in the early MS group except that the PPV of NEDA in the early MS group peaked sooner at 2 years. No significant difference was found between the early and established MS cohorts in the proportion who had NEDA annually during the 7-year study period (Table 2). Regarding the individual components of disease activity, there was a significant difference between the early and established cohorts in terms of the absence of MRI disease activity at 2 to 7 years, with the results favoring the established MS cohort in each instance. In Table 3, we provide NEDA status as reported in published studies compared with our findings in the CLIMB cohort.

With multiple therapies for relapsing forms of MS, NEDA has become a treatment goal and outcome measure. In the present era of disease-modifying therapy, it is uncommon for patients with MS to become significantly disabled in the first 5 years of disease. This occurrence has made it difficult to prognosticate early in the disease course and adapt treatment strategies accordingly. NEDA has garnered increasing attention as a measure that may allow for earlier and more accurate prognostication and has become a secondary outcome measure in clinical trials for new disease-modifying therapies in MS.³ However, its persistence over time and its predictive power for long-term prognosis are unknown. The CLIMB study provided a unique opportunity to examine NEDA over time in a real-world cohort that includes clinical evaluation every 6 months and yearly MRIs. We found that patients continued to lose NEDA status during the 7-year period, although the slope decreased with time. Of interest, NEDA at 2 years predicted disability at 7 years nearly as well as NEDA at 5 years.

The relevance of early changes in clinical or MRI measures with respect to long-term outcomes has been reported,^16-20 including the finding that early MRI disease activity while on interferon therapy predicts poor long-term outcomes.^21,22 The combined predictive power of early clinical and MRI changes for disability prognosis has not been as extensively studied. For example, one study²³ reported that a composite measure of 1-year MRI lesion activity and relapses accounted for most of the change in EDSS scores observed at 2 years. In another study²⁴ using a composite score that integrated new T2 lesions and relapses, disease activity in an interferon clinical trial predicted disability progression during the subsequent 3 years. Our findings suggest that the combined NEDA measure allows for better early prediction of freedom from progression at long-term follow-up (7 years) than absence of relapses, disability progression, relapses and disability progression, or new MRI lesions alone.

Previous reports^5-10,25 on NEDA or disease-free status have evaluated patients receiving a single disease-modifying therapy, often as post hoc analysis of clinical trials. As opposed to clinical trial data, our 7-year CLIMB cohort reflects a real-world cohort of patients with RRMS. The average patient in our study underwent disease-modifying therapy for approximately 75% of the 7-year study period, which is consistent with adherence rates reported in the literature.²⁶Quiz Ref IDMost patients used first-line injectable agents because they were enrolled in 2000 through 2005 before oral disease-modifying therapies and natalizumab became available. Although comparing drug performance on NEDA is of great potential interest, such comparisons may be of limited value in an observational setting because of patient selection into specific treatment groups based on antecedent disease severity. Due to the small number of patients in each treatment group at each time point, we were underpowered to determine how specific therapies affect the predictive value of NEDA. This important question may be addressed in future studies with a larger number of patients.

We observed that performance on NEDA was similar in early and established MS during the entire study period. This finding may be related to the incorporation of inflammatory outcomes (relapses, new MRI lesions) and sustained progression (which may be more related to neurodegeneration) into the NEDA composite measure. Patients with secondary progressive MS were not part of our study, and their exclusion may account for the similarities in individual component measures across the groups. Moreover, although older age and longer disease duration are typically associated with a worse prognosis in MS, the exclusion of patients with secondary progressive MS may have selected for an older population with relatively more benign disease, which may explain why patients with NEDA were more likely to be slightly older with longer disease duration.

Although NEDA status may strongly predict good long-term outcomes, loss of NEDA status is not necessarily a poor prognostic sign. Clinical experience indicates that patients may have a relapse or new MRI lesion and do well in the long term, and the low NPV of NEDA in our study is consistent with this. Studies^27,28 have suggested that inflammatory activity in MS does not always correlate with the rate of later disability progression.

The dissociation between MRI and clinical disease activity that we observed has been noted in other natural history studies^29,30 and may in part be due to the shortcomings of our current imaging tools. Although 30.6% to 42.9% of our cohort had no evidence of disease activity by either imaging or clinical measures at each annual time point, they had disease activity by the other measure. We found that during the long-term follow-up a greater proportion of patients lose NEDA status based on clinical rather than MRI criteria, a finding that was somewhat surprising given that MRI events were primarily responsible for loss of NEDA in previous investigations of shorter duration.¹

Other MRI measures besides new T2 or T1 gadolinium-enhancing lesions may prove to be more informative markers of disease activity. There is evidence that whole-brain atrophy, gray matter atrophy, and T1 hypointense lesions have stronger correlations with long-term disability than T2 or gadolinium-enhancing lesions.^31-36 To incorporate these more advanced MRI measures into NEDA, the threshold for change in each measure would need to be established. Moreover, such measures are not yet routinely assessed in clinical practice. Furthermore, spinal cord imaging is important to assess the true rate of radiographic disease activity. We observed that the addition of spinal cord imaging to the assessment of MRI activity in our cohort led to additional losses of 7% to 11% annually in the number of patients free of MRI disease activity.

The T25FW is an easy-to-obtain outcome measure that may offer further insights into disease activity. Early changes in the T25FW may be more sensitive than the EDSS for predicting long-term disability in progressive MS.³⁷ If change in the T25FW can be demonstrated to be a sensitive and reproducible outcome measure that is correlated with long-term disability status, it could be incorporated into NEDA as well.

Quiz Ref IDThe ultimate goal of MS therapy is to treat patients so as to prevent disability. It would seem logical that achieving such a goal will be linked to treating in a fashion such that there is no evidence of disease activity by clinical and MRI measures. Our finding that NEDA at 2 years, as measured with annual imaging and biannual clinical assessment, can predict disability at 7 years requires validation in other cohorts and additional longitudinal follow-up. The probability of developing progressive disease might decrease after 15 years, which may be an ideal target for assessment of NEDA and its predictive power.^38,39 Our ongoing 2200-patient CLIMB study provides an ideal cohort in whom to investigate NEDA in the current era of MS therapy. Although NEDA has the potential to become not only a key outcome measure of disease-modifying therapy but also a treat-to-target goal, it will require a comprehensive approach that integrates advances in MRI technology, linkage of blood and cerebrospinal fluid biomarkers, and a high degree of cooperation among investigators.

Accepted for Publication: October 2, 2014.

Corresponding Author: Howard L. Weiner, MD, Partners Multiple Sclerosis Center, Brigham and Women’s Hospital, 1 Brookline Pl, Ste 227, Brookline, MA 02445 (hweiner@rics.bwh.harvard.edu).

Published Online: December 22, 2014. doi:10.1001/jamaneurol.2014.3537.

Author Contributions: Dr Rotstein had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Rotstein, Weiner.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Rotstein.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Rotstein, Healy.

Obtained funding: Weiner.

Administrative, technical, or material support: Malik, Chitnis, Weiner.

Study supervision: Chitnis, Weiner.

Conflict of Interest Disclosures: Dr Rotstein reported receiving research support from the Multiple Sclerosis Society of Canada and serving as a consultant for sanofi-aventis. Dr Healy reported receiving grant support from Merck Serono and Novartis. Dr Malik reported receiving grant support from Merck Serono. Dr Chitnis reported serving as a consultant for Biogen-Idec, Novartis, and Alexion and receiving grant support from EMD Serono and Novartis. Dr Weiner reported receiving personal compensation for consulting and speaking activities and serving on the scientific advisory board of companies, including Biogen Idec, Novartis, EMD Serono, Teva Neurosciences, GSK, Nasvax, Xenoport, and Genzyme. He also reported receiving grant support from EMD Serono. No other disclosures were reported.

Funding/Support: Merck Serono provided partial financial support for data collection and reviewed the manuscript.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, or approval of the manuscript; and the decision to submit the manuscript for publication.